Abstract: In this paper, I theoretically and experimentally compare an employer's profit from two contest designs: a standard winner-take-all tournament with a single prize, and two parallel tournaments with different prizes. I develop a simple model that illustrates the comparative statistics affecting the relative performance of both systems and conduct a set of experiments to inspect these predictions in the lab. I find three main results from the experiments. First, as predicted by the theory, the employer's profit is higher under the parallel tournament design when the agents' abilities differ greatly. Second, the employer's profit is higher under the single-prize tournament when agents' abilities are similar. Third, high-ability agents under-participate and low-ability agents over-participate in the high-prize tournament.

Abstract: This paper studies a two-stage decentralized matching game with time discounting. In the game, firms sequentially make directed offers to workers. I assume that a firm, which ranks a worker as the highest, is ranked in the bottom in that worker's preference list and there is no conflict of interest among agents on the same side of the market. When agents have complete information and the time discount is sufficiently high or sufficiently low, I find that the unique sub-game Nash equilibrium outcome of the game yields the firm-optimal stable match. However, when the time discount is between these threshold levels, the unique sub-game Nash equilibrium outcome of the game yields the worker-optimal stable match.